1 Project Summary

The goal of this study is to characterize the use of oral antibiotics for the treatment of acne in patients aged 10-45. We conducted a retrospective, observational, cohort study using the Observational Health Data Sciences and Informatics (OHDSI) database and data from Health Data Compass at the University of Colorado.

Eligibility of those included is displayed below.

2 Summary Statistics

Continuous variables are summarized by mean, standard deviation (SD), median, minimum and maximum whereas categorical variables are presented as the number in each group and percent of column total. Table 2.1 compares patients with and without an antibiotic treatment course of interest and Table 2.2 compares patients with and without an isotrentinoin prescription. Comparisons between groups are conducted using either Chi square tests for proportions for categorical variables (sex, race, ethnicity) or t-tests for continuous variables (age). All tests were evaluated at the 0.05 significance level. R version 3.6.0 was used for all summaries and figures.

Overall, the patient population is an overwhelmingingly female composed of mostly non-hispanic white individuals with a median age of 20. Demographics of those who recieved antiobiotics and of those who recieved isotretinoin treatment courses are overall comparable.

2.1 Characteristics of Patients with and without an Antibiotic Prescription for Acne

User Prevalence None
(N=10565)
Recieved Medication
(N=4756)
Overall
(N=15321)
Age
Mean (SD) 22.9 (8.56) 21.0 (8.41) 22.3 (8.56)
Median [Min, Max] 21.0 [10.0, 45.0] 18.0 [10.0, 45.0] 20.0 [10.0, 45.0]
Sex
Female 7822 (74.0%) 2995 (63.0%) 10817 (70.6%)
Male 2738 (25.9%) 1760 (37.0%) 4498 (29.4%)
Missing 5 (0.0%) 1 (0.0%) 6 (0.0%)
Race
American Indian or Alaska Native 38 (0.4%) 15 (0.3%) 53 (0.3%)
Asian 295 (2.8%) 106 (2.2%) 401 (2.6%)
Black or African American 285 (2.7%) 91 (1.9%) 376 (2.5%)
Native Hawaiian or Other Pacific Islander 17 (0.2%) 8 (0.2%) 25 (0.2%)
White 8541 (80.8%) 3926 (82.5%) 12467 (81.4%)
Missing 1389 (13.1%) 610 (12.8%) 1999 (13.0%)
Ethnicity
Hispanic or Latino 899 (8.5%) 403 (8.5%) 1302 (8.5%)
Not Hispanic or Latino 9022 (85.4%) 3972 (83.5%) 12994 (84.8%)
Missing 644 (6.1%) 381 (8.0%) 1025 (6.7%)
Diagnosis
Acne 10534 (99.7%) 4740 (99.7%) 15274 (99.7%)
Acne conglobata 10 (0.1%) 2 (0.0%) 12 (0.1%)
Acne varioliformis 8 (0.1%) 6 (0.1%) 14 (0.1%)
Excoriated acne 11 (0.1%) 7 (0.1%) 18 (0.1%)
Tropical acne 2 (0.0%) 1 (0.0%) 3 (0.0%)

2.2 Characteristics of Patients with and without Isotretinoin Prescription

User Prevalence None
(N=13217)
Recieved Medication
(N=2104)
Overall
(N=15321)
Age
Mean (SD) 22.7 (8.75) 19.6 (6.64) 22.3 (8.56)
Median [Min, Max] 21.0 [10.0, 45.0] 18.0 [10.0, 45.0] 20.0 [10.0, 45.0]
Sex
Female 9685 (73.3%) 1132 (53.8%) 10817 (70.6%)
Male 3527 (26.7%) 971 (46.2%) 4498 (29.4%)
Missing 5 (0.0%) 1 (0.0%) 6 (0.0%)
Race
American Indian or Alaska Native 50 (0.4%) 3 (0.1%) 53 (0.3%)
Asian 361 (2.7%) 40 (1.9%) 401 (2.6%)
Black or African American 342 (2.6%) 34 (1.6%) 376 (2.5%)
Native Hawaiian or Other Pacific Islander 23 (0.2%) 2 (0.1%) 25 (0.2%)
White 10736 (81.2%) 1731 (82.3%) 12467 (81.4%)
Missing 1705 (12.9%) 294 (14.0%) 1999 (13.0%)
Ethnicity
Hispanic or Latino 1150 (8.7%) 152 (7.2%) 1302 (8.5%)
Not Hispanic or Latino 11239 (85.0%) 1755 (83.4%) 12994 (84.8%)
Missing 828 (6.3%) 197 (9.4%) 1025 (6.7%)
Diagnosis
Acne 13178 (99.7%) 2096 (99.6%) 15274 (99.7%)
Acne conglobata 7 (0.1%) 5 (0.2%) 12 (0.1%)
Acne varioliformis 12 (0.1%) 2 (0.1%) 14 (0.1%)
Excoriated acne 17 (0.1%) 1 (0.0%) 18 (0.1%)
Tropical acne 3 (0.0%) 0 (0%) 3 (0.0%)

3 Results

When comparing those who recieved an antibiotic course of interest to those who did not among variables of interest, age, sex, and race were significantly different. Due to large size of this large cohort, very small differences will reach statistical significance even though the differences are not clinically relevent. In this study, age and race are significantly different but the magnitude of the effect is very small. In contrast to age and race, sex appeared to have clinically meaningful results with an increased likelihood of recieveing medication among male patients despite an overwhelmingly female population (70.6% female). Among the 4,605 male patients in the eligible cohort, 1,810 (39%) of them recieved an antibioitic treatment of interest in comparison to 3,109 (28%) of the 11,093 eligible women. This supports the hypothesis that while more female patients seek treatment for acne, they are prescribed less often than male patients as male patients are hypothesized to only seek treatment following a more severe acne diagnosis.

There was a similar trend for patients using isotretinoin. Age, sex, and race all produce significant results but only sex appears to have a clinically meaningful difference between patients with and without an isotretinoin treatment. Among the 4605 eligible men, 988 recieved isotretinoin of the eligible (21.5%) compared to 1159 of 11,093 women (10.4%).

Among those who did get some level of antibiotic treatment, the mean amount of time between first acne diagnoses and drug exposure start date was less than one day.

4 Visualizations

4.1 Prescription Count

Minocycline was the most commonly prescribed antibiotic of interest followed by Doxycycline.

4.2 Prescription Count over Time

allows for multiple prescriptions per person

allows for multiple prescriptions per person

4.3 Isotrentinoin Medications

A vastly female cohort

5 Isotretinoin Prescriptions Over Time

allows for multiple prescriptions per person

5.1 Diagnoses Count

A diagnoses of “Acne” was by far the most diagnosesd in the eligible population.

5.2 How many visits resulted in prescriptions

By far, most individuals visited the doctor and were not prescripted any medication (10,784). About a quarterof that (n = 2,527, 23%) of eligible individuals had 1 visit that resulted in medication.

5.3 Which medications Reoccur?

Doxycycline and Minocyline are by far the most prescriped medications with minocycline the most commonly re-prescribed to an indivudal.

Same plot pictured below but focusing on Doxycycline and Minocycline and collapsing Azithromycin, Amoxicillin, Cephalexin, & Sufamethoxazole.

6 Appendix

There are 7 affiliated csv files; 6 for antibiotics of interest and 1 for Isotretinoin. Eligible prescription drug concept IDs are listed in the “Id” column. Minimum length for all drug courses was a quantity of 27 unless otherwise listed in the “Standards” csv file. All graphs depicting multiple prescriptions per person were taken from multiple occurences in the dataset and potentially excluding refills.